Everything Freedom Tales writes back to you is generated by a language model, in real time, in response to what you type. There is no author reviewing it before it reaches you, and no fixed script it is reading from.
The setting, the characters and the situation you begin in are authored in advance by a human. What happens once you start playing — every line of narration, every piece of dialogue, every detail a character reveals — is written by the model as it goes, based on what you say and everything the story has recorded so far.
The model is instructed to keep track of what it has already established and stay consistent with it, and to let characters be genuinely guarded about what they don't want to say. It can still make mistakes: it can occasionally contradict something said earlier, or narrate something that doesn't quite fit. That is a property of the technology, not a hidden human choice.
Freedom Tales is not a chat assistant, and the characters in it are not real people, past or present, regardless of what a story might imply about who they resemble. Nothing a character says is a factual claim, an opinion held by us, or advice.
We use models from Anthropic and, for some scenarios, other providers reached through OpenRouter. Which model narrated a given story is not shown in the story itself, to keep the fiction the point rather than the technology.